Innovation Economics · Sankey

Where the AI capital actually goes.

Trailing twelve months of disclosed AI funding, mapped from source region to destination bucket. The ribbon widths scale to dollar volume. Three patterns are immediately legible: US dominance in foundation models, China's parity in infrastructure, and SSA's near-absence from frontier compute.

United States → Foundation models: $38.5B United States → Infrastructure & compute: $27.2B China → Infrastructure & compute: $19.8B United States → Applications: $18.9B China → Foundation models: $14.2B China → Public-sector AI: $12.1B Europe → Applications: $11.3B Europe → Infrastructure & compute: $9.2B China → Applications: $8.7B Europe → Foundation models: $7.8B MEA → Infrastructure & compute: $6.8B United States → Research labs: $6.4B Europe → Public-sector AI: $5.6B China → Research labs: $4.9B United States → Public-sector AI: $4.1B Europe → Research labs: $3.4B MEA → Public-sector AI: $3.2B MEA → Foundation models: $1.9B MEA → Applications: $1.2B Sub-Sah. Africa → Applications: $850M MEA → Research labs: $600M Sub-Sah. Africa → Infrastructure & compute: $320M Sub-Sah. Africa → Public-sector AI: $230M Sub-Sah. Africa → Research labs: $110M Sub-Sah. Africa → Foundation models: $90M United States $95.1B Europe $37.3B China $59.7B MEA $13.7B Sub-Sah. Africa $1.6B Foundation models $62.5B Infrastructure & compute $63.3B Applications $41.0B Research labs $15.4B Public-sector AI $25.2B
Source: Briter Bridges + Partech Africa rolling compilation, AI deal disclosures. Last verified .

Total flow

$207.4B

Disclosed AI venture & growth capital, trailing twelve months.

US foundation-model share

19%

Single ribbon, US → foundation models, against the global total.

SSA share of frontier infra

0.51%

SSA ribbon into infrastructure as a share of all infrastructure capital.

← All models